Recent activities in attacks on automated teller machines have shown a sophistication that has grown to a degree, where it is not always technically possible to prevent the attack. This paper describes an approach for anomaly and attack detection for ATMs. The approach works on multiple levels. First, we use sensor fusion on the low-level hardware sensors to get robust information about the device state. Second, we use a new model-based and self-learning anomaly detection method on the diagnosis data of all ATM devices to robustly detect anomalies in the system that might indicate an attack on the machine.
CITATION STYLE
Perner, P. (Ed.). (2010). Advances in Data Mining. Applications and Theoretical Aspects (Vol. 6171, pp. 405–417). Berlin, Heidelberg: Springer Berlin Heidelberg. https://doi.org/10.1007/978-3-642-14400-4
Mendeley helps you to discover research relevant for your work.